We have employed a novel strategy to identify molecular targets for the therapy and prevention of common cancers, with an initial focus on colon cancer. One of the major limitations of chemotherapeutic agents is the relatively narrow therapeutic: toxic ratio that results from the commonality of the therapeutic target in both the cancerous and normal tissues. Therefore, we sought to identify potential therapeutic targets that were relatively cancer specific by using gene expression information available in public databases coupled with the use of high-throughput gene expression array analysis. Using the Digital Gene expression Displayer to mine expression libraries available through the NCI's Cancer Genome Anatomy Project, we identified 665 sequences that were expressed in colon cancer libraries but not in essential normal tissues. Of these, 356 were represented on Affymetrix expression arrays and this allowed for a comparison of the expression of these genes in colon cancer tissues and cell lines, and essential normal tissues. This investigation coupled with subsequent confirmation using quantitative RT-PCR resulted in the identification of two previously uncharacterized genes whose expression was relatively specific to a substantial proportion of colon cancer samples tested. We are currently pursuing "knock out" of these genes using small interfering RNA (siRNA). These assays will be used to verify the role each may play in maintenance of the malignant phenotype. The approach of using available databases coupled with high-throughput technologies for the identification of relatively cancer selective gene targets may constitute a rapid and efficient method for the development of active and less toxic chemotherapeutic agents. The identification of these genes as potential targets was based entirely on their expression profile rather than dependent on prior knowledge of their intracellular function(s). In addition to providing targets of interest for drug development, such targets could also be exploited for drug delivery, as molecular markers of disease, or as surrogate markers for drug delivery and drug effectiveness.